Breaking windows???????
Don't you mean with pois chiches it will be breaking wind????
Ciao,
Jocko
Don't you mean with pois chiches it will be breaking wind????
Ciao,
Jocko
local stuff
Koinichiwa,
I have a few comments on your last post:
I am at a loss why you advocate using EXACTLY the same opamp for supply and as amplifier. You stated in your earlier post that the supply opamp bandwidth should be larger than the amp opamp itself. You can do that by using identical opamps but it seems to be more sensible to chose opamps optimal for the application, for instance a low THD one for the amp and a wideband one for the supply.
I can in general agree to your noise/ripple calculation, but please note that PSRR and noise specs are normally referred to the input. With a gain of 10, all numbers get worse by 20dB, which you didn't address.
On the signal current modulation, these in class AB have of course the shape of rectified sine waves. I don't know if I would call that benign. It is harmonic, but I suspect there are some nasty odd ones which may be audibly objectionably. But it's all a matter of level as well, of course. In class A, the signal supply modulation would be purely sinussoidal (with a sine signal) and this may be one more factor in the advantage of the class A amp.
Jan Didden
Koinichiwa,
I have a few comments on your last post:
I am at a loss why you advocate using EXACTLY the same opamp for supply and as amplifier. You stated in your earlier post that the supply opamp bandwidth should be larger than the amp opamp itself. You can do that by using identical opamps but it seems to be more sensible to chose opamps optimal for the application, for instance a low THD one for the amp and a wideband one for the supply.
I can in general agree to your noise/ripple calculation, but please note that PSRR and noise specs are normally referred to the input. With a gain of 10, all numbers get worse by 20dB, which you didn't address.
On the signal current modulation, these in class AB have of course the shape of rectified sine waves. I don't know if I would call that benign. It is harmonic, but I suspect there are some nasty odd ones which may be audibly objectionably. But it's all a matter of level as well, of course. In class A, the signal supply modulation would be purely sinussoidal (with a sine signal) and this may be one more factor in the advantage of the class A amp.
Jan Didden
Thorsten:
>Reject this by the 120db PSRR of the same Op-Amp<
Many opamps have far less than 120dB PSRR. And PSRR typically describes a frequency-dependent curve, so I fail to see the point in mentioning a specific number unless the measurement frequency is also noted. Usually, the higher the measurement frequency, the worse the PSRR.
And as Jan noted, the PSRR typically decreases in direct proportion to the amount of gain that the opamp is asked to provide.
regards, jonathan carr
>Reject this by the 120db PSRR of the same Op-Amp<
Many opamps have far less than 120dB PSRR. And PSRR typically describes a frequency-dependent curve, so I fail to see the point in mentioning a specific number unless the measurement frequency is also noted. Usually, the higher the measurement frequency, the worse the PSRR.
And as Jan noted, the PSRR typically decreases in direct proportion to the amount of gain that the opamp is asked to provide.
regards, jonathan carr
local stuff
Indeed Jonathan,
The PSRR's I mentioned in my "thinking out loud" post were the minimum specified up to 20kHz or so for the opamp I had in mind, as a kind of worst case. At 1kHz they may approach 80 or even 90dB, but I have never seen 120dB either.
Jan Didden
Indeed Jonathan,
The PSRR's I mentioned in my "thinking out loud" post were the minimum specified up to 20kHz or so for the opamp I had in mind, as a kind of worst case. At 1kHz they may approach 80 or even 90dB, but I have never seen 120dB either.
Jan Didden
That's a lot of dB............
From the 120 dB guy........
"Unlike you I have worked much with the various regulator topologies. I find that there are both objective (measured) and subjective (sound) differences. Maybe you would wish to actually empirically evaluate the subject instead of just arguning from a theoretical viewpoint that is by far to abstract and simplified to be usable in reality? "
120 dB is a change in output of 1 microvolt for every volt change at the supply. In theory one might get that number with a very good regulator design, but in real life its a little harder. It is real challenge to measure by the way, for even a techno nerd with very good lab. Walt Jung has written about making measurements at this level and even he said that it was no picnic even with excellent lab equipment. My hat's off to the 120dB guys measuring skills, or perhaps one can assume that you can get this level of performance just because the data sheet says so. That is if you actually can actually interpret data that's there, like seeing the fact that it changes by orders of magnitude over the audio band width. I sure wish circuits would work as good the data sheet would have you believe. One interesting thing is that they often measure the parts under the conditions where the part gives the best performance. How many of you are running your three terminal regulators at 0.5 amps DC load current where the output impedance curves are usually shown? Don't forget that a data sheet is intended as an advertising vehicle to show the part in its best light so you will design it in and your company will buy lots of them. Like with all advertisements it pays to read the fine print very carefully.
Empirically yours,
Fred
/has actually built and listened to power supplies on a few ocassions and can even read data sheets.
From the 120 dB guy........
"Unlike you I have worked much with the various regulator topologies. I find that there are both objective (measured) and subjective (sound) differences. Maybe you would wish to actually empirically evaluate the subject instead of just arguning from a theoretical viewpoint that is by far to abstract and simplified to be usable in reality? "
120 dB is a change in output of 1 microvolt for every volt change at the supply. In theory one might get that number with a very good regulator design, but in real life its a little harder. It is real challenge to measure by the way, for even a techno nerd with very good lab. Walt Jung has written about making measurements at this level and even he said that it was no picnic even with excellent lab equipment. My hat's off to the 120dB guys measuring skills, or perhaps one can assume that you can get this level of performance just because the data sheet says so. That is if you actually can actually interpret data that's there, like seeing the fact that it changes by orders of magnitude over the audio band width. I sure wish circuits would work as good the data sheet would have you believe. One interesting thing is that they often measure the parts under the conditions where the part gives the best performance. How many of you are running your three terminal regulators at 0.5 amps DC load current where the output impedance curves are usually shown? Don't forget that a data sheet is intended as an advertising vehicle to show the part in its best light so you will design it in and your company will buy lots of them. Like with all advertisements it pays to read the fine print very carefully.
Empirically yours,
Fred
/has actually built and listened to power supplies on a few ocassions and can even read data sheets.
Re: local stuff
Koinichiwa,
I did not ADVOCATE this, I was merely playing a number game. In this case I I was pointing out that the calculation where based on so to speak "cascading" the same Op-Amp.
Sure. I completely agree.
I know. I seem to have missed that you where using the OPA at 20db gain. The Regs in my original circuit had an AC and DC gain of 2, but being CFB types that does little to the PSRR (bandwidth etc.).
They should only appear at high levels, assuming the design is sufficiently competent. Most audio circuitry should be in Class A most of the way, really.
Sayonara
Koinichiwa,
janneman said:
I did not ADVOCATE this, I was merely playing a number game. In this case I I was pointing out that the calculation where based on so to speak "cascading" the same Op-Amp.
janneman said:
Sure. I completely agree.
janneman said:
I know. I seem to have missed that you where using the OPA at 20db gain. The Regs in my original circuit had an AC and DC gain of 2, but being CFB types that does little to the PSRR (bandwidth etc.).
janneman said:
They should only appear at high levels, assuming the design is sufficiently competent. Most audio circuitry should be in Class A most of the way, really.
Sayonara
Koinichiwa Carr San,
Please DO NOT quote me out of context. Please quote the full passage:
"Reject this by the 120db PSRR of the same Op-Amp with 60db PSRR used as regulator and as amplifier"
This implies that the 120db rejection are achieved by the 60db rejection of the actual regulator circuit, whose output noise (in case of the same Op-Amp as the amplifier being used) is being rejected by 60db by the Amplifier circuit. The endresult, any noise of the preregulated supply is rejected by 120db.
The LM6181/82 suggested in my particular implementation of such an approach has a PSRR of > 60db past 20KHz, so a "system" PSRR of 120db out to 20KHz is not much or a problem.
As stated, the actual noise limits of such circuit have more to do with reference design and reducing the reference's noise.
Sayonara.
jcarr said:
Please DO NOT quote me out of context. Please quote the full passage:
"Reject this by the 120db PSRR of the same Op-Amp with 60db PSRR used as regulator and as amplifier"
This implies that the 120db rejection are achieved by the 60db rejection of the actual regulator circuit, whose output noise (in case of the same Op-Amp as the amplifier being used) is being rejected by 60db by the Amplifier circuit. The endresult, any noise of the preregulated supply is rejected by 120db.
The LM6181/82 suggested in my particular implementation of such an approach has a PSRR of > 60db past 20KHz, so a "system" PSRR of 120db out to 20KHz is not much or a problem.
As stated, the actual noise limits of such circuit have more to do with reference design and reducing the reference's noise.
Sayonara.
Honorable LM6181 regulator-san is number 10
"The LM6181/82 suggested in my particular implementation of such an approach has a PSRR of > 60db past 20KHz, so a "system" PSRR of 120db out to 20KHz is not much or a problem."
Yes and unfortunately it has the same PSRR at 120 Hz, where the LM317 is about 4 or 5 times better. 10 times better with a preloaded DC output load of a few 100 mAs. The LM6181 output impedance is about a quarter of an ohm in series with 40 to 50 nH of inductance. This is about 100 times higher than a LM317 in the 100 to 1 kHz region. The LM6181 app note shows pretty severe overshoot and ringing with less than a 10 pF load. (where the cable reflections you expressed concerns about, are)
At DC loads over about 10 mA your op amp push pull regulators output stage is running single ended by the way and it is not ush pull any more. I'll let Jocko tell you what he thinks about the onics for current feedback op amps. The last one I used was the AD 846 and it took a resistor to the compensation terminal to sound anywhere close to decent.
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=37994
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=39110
"As stated, the actual noise limits of such circuit have more to do with reference design and reducing the reference's noise."
Absolutely, why not try it? I don't think the LM431 is as near as good as the LM 329 whose noise spec is characterized on the Data sheet as opposed to "Low output noise voltage" promise for the LM431 on its data sheet. To make matters worse the noise filter corner is set by the 0.15 output impedance of the
LM431 and the 10 uF cap at about 100 kHz. The noise is also mutiplied by the op amps closed loop gain. (Yes there is a way to put a large cap in the feedback loop to drop the AC noise gain to unity with DC gain still greater than one, and have still use a current feedback op amp) Actually the feedback for the LM431 is falling of and the inductive output impedance is resonating with the high Q low ESR 10 uF Oscon cap. It is encouraging to see your enthusism for low reference noise circuits, but what the heck happened to that goal on this design?
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=39029
I am not surpassed that you gave this up for tubes and passively filtered supplies. You might want fall back on your subjective skills when it comes to designing and building power supplies.
Anybody that wants to take a shot at this circuit should have about a 500 MHz scope and resistive scope probe. A regular scope probes capacitance will alter the stability of this circuit when measuring it. I have seen stuff like this work when you put a probe on it and act flaky the when probe is removed. It makes troubleshooting a real joy for the unwary, and I speak from personal experience (digital and analog circuits) on that one.
I hope this was empirical enough for everyone. I wonder why people didn't like me to come to their telecom design reviews. I had to be nice (it was tough) during the functional equivalent of telling a mother that was the ugliest baby I had ever seen! I usually snuck in the back after they started.
"The LM6181/82 suggested in my particular implementation of such an approach has a PSRR of > 60db past 20KHz, so a "system" PSRR of 120db out to 20KHz is not much or a problem."
Yes and unfortunately it has the same PSRR at 120 Hz, where the LM317 is about 4 or 5 times better. 10 times better with a preloaded DC output load of a few 100 mAs. The LM6181 output impedance is about a quarter of an ohm in series with 40 to 50 nH of inductance. This is about 100 times higher than a LM317 in the 100 to 1 kHz region. The LM6181 app note shows pretty severe overshoot and ringing with less than a 10 pF load. (where the cable reflections you expressed concerns about, are)
At DC loads over about 10 mA your op amp push pull regulators output stage is running single ended by the way and it is not ush pull any more. I'll let Jocko tell you what he thinks about the onics for current feedback op amps. The last one I used was the AD 846 and it took a resistor to the compensation terminal to sound anywhere close to decent.
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=37994
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=39110
"As stated, the actual noise limits of such circuit have more to do with reference design and reducing the reference's noise."
Absolutely, why not try it? I don't think the LM431 is as near as good as the LM 329 whose noise spec is characterized on the Data sheet as opposed to "Low output noise voltage" promise for the LM431 on its data sheet. To make matters worse the noise filter corner is set by the 0.15 output impedance of the
LM431 and the 10 uF cap at about 100 kHz. The noise is also mutiplied by the op amps closed loop gain. (Yes there is a way to put a large cap in the feedback loop to drop the AC noise gain to unity with DC gain still greater than one, and have still use a current feedback op amp) Actually the feedback for the LM431 is falling of and the inductive output impedance is resonating with the high Q low ESR 10 uF Oscon cap. It is encouraging to see your enthusism for low reference noise circuits, but what the heck happened to that goal on this design?
http://katalogi.iele.polsl.gliwice.pl/en/download.php?id=39029
I am not surpassed that you gave this up for tubes and passively filtered supplies. You might want fall back on your subjective skills when it comes to designing and building power supplies.
Anybody that wants to take a shot at this circuit should have about a 500 MHz scope and resistive scope probe. A regular scope probes capacitance will alter the stability of this circuit when measuring it. I have seen stuff like this work when you put a probe on it and act flaky the when probe is removed. It makes troubleshooting a real joy for the unwary, and I speak from personal experience (digital and analog circuits) on that one.
I hope this was empirical enough for everyone. I wonder why people didn't like me to come to their telecom design reviews. I had to be nice (it was tough) during the functional equivalent of telling a mother that was the ugliest baby I had ever seen! I usually snuck in the back after they started.
Attachments
Hi,
Now I know why I like tube electronics so much...
Even the most sophisticated reg is a cinch and works and works and works....
I wonder why you guys make your life so complicated....it's short enough as it is.
And does all of the previous post make for better sound?
I wonder,😉
Now I know why I like tube electronics so much...
Even the most sophisticated reg is a cinch and works and works and works....
I wonder why you guys make your life so complicated....it's short enough as it is.
And does all of the previous post make for better sound?
I wonder,😉
fly me to the moon.
"We choose to design Solid State audio in this decade and do the other things, not because they are easy but because they are hard, because that goal will serve to organize and measure the best of our energies and our skills ... we do not know what benefits await us ... [but] music is there and we are going to find it."
"Some people see transistor circuits as they are and say why. I dream transistor circuits that never were and say why not?"
"We choose to design Solid State audio in this decade and do the other things, not because they are easy but because they are hard, because that goal will serve to organize and measure the best of our energies and our skills ... we do not know what benefits await us ... [but] music is there and we are going to find it."
"Some people see transistor circuits as they are and say why. I dream transistor circuits that never were and say why not?"
Attachments
Re: Honorable LM6181 regulator-san is number 10
Koinichiwa Diekman San,
As a pre-regulated supply is supplied the 120Hz rejection is not an issue.
Except, the cable capacity as such does not directly load the regulators, merely indirectly.
While all you say is true you are as usual of poor reading ability, as you seek not to understand but merely to "proove a point". Had you bothered reading yould have noticed in my original post the following remark:
"PS, attached a rendering of an experimental linestage from a few years ago, drawn by Carlos M. from a description by me, no guarantee for accuracy, I never drew a schema of this, all this is from memory and my memory is notoriously bad.
I think I had a 1k resistor (maybe 10k?) between reference Voltage and Op-Amp input for the reg."
With 10k and probably around 47uF Os-Con (that's what i tend to keep around in my parts box) we have a 1st order filter with 0.33Hz Corner Frequency, with 1k (I really can't remember - I probably used 1k) we have a 3.3Hz corner.
This means at 33Hz our noise is aready 20db down from that of the bare LM431, at 330Hz it's 40db down. And I agree, even so the noiselevels could be improved. Of course, I was not actually chasing lowest noise myself (Didden San is), I was chasing a very different parameter.
I know. Place the resistor needed for stability in series with the inverting input. As said, lowest noise and lowest output impedance as such where not design goals.
Fully agreed. Don't even bother to try to measure the PSU Lines with a normal 'scope. The regulator and amplifier circuit will oscillate with the load capacitance of a 'scope probe, while I found it stable in normal operation. For the Amplifier, we have a large enough value "build out" resistor to make the circuit operationally unconditionally stable, except when someone pokes around with a 'scope.
Funny. I seem to have missed you actually building the circuit. So empiricism seems not to have featured. But semantics is hardly a specialty of mine.
Sayonara
Koinichiwa Diekman San,
Fred Dieckmann said:
As a pre-regulated supply is supplied the 120Hz rejection is not an issue.
Fred Dieckmann said:
Except, the cable capacity as such does not directly load the regulators, merely indirectly.
Fred Dieckmann said:
While all you say is true you are as usual of poor reading ability, as you seek not to understand but merely to "proove a point". Had you bothered reading yould have noticed in my original post the following remark:
"PS, attached a rendering of an experimental linestage from a few years ago, drawn by Carlos M. from a description by me, no guarantee for accuracy, I never drew a schema of this, all this is from memory and my memory is notoriously bad.
I think I had a 1k resistor (maybe 10k?) between reference Voltage and Op-Amp input for the reg."
With 10k and probably around 47uF Os-Con (that's what i tend to keep around in my parts box) we have a 1st order filter with 0.33Hz Corner Frequency, with 1k (I really can't remember - I probably used 1k) we have a 3.3Hz corner.
This means at 33Hz our noise is aready 20db down from that of the bare LM431, at 330Hz it's 40db down. And I agree, even so the noiselevels could be improved. Of course, I was not actually chasing lowest noise myself (Didden San is), I was chasing a very different parameter.
Fred Dieckmann said:
I know. Place the resistor needed for stability in series with the inverting input. As said, lowest noise and lowest output impedance as such where not design goals.
Fred Dieckmann said:
Fully agreed. Don't even bother to try to measure the PSU Lines with a normal 'scope. The regulator and amplifier circuit will oscillate with the load capacitance of a 'scope probe, while I found it stable in normal operation. For the Amplifier, we have a large enough value "build out" resistor to make the circuit operationally unconditionally stable, except when someone pokes around with a 'scope.
Fred Dieckmann said:
Funny. I seem to have missed you actually building the circuit. So empiricism seems not to have featured. But semantics is hardly a specialty of mine.
Sayonara
I have been eyeing this 6181 preamp for quite some time now and I would love to be the guineau pig to build it and report back to this thread. We could even feed some inverted gainclones with it. Thorsten? Batteries or mains supply?
Regulators Rudux
"As a pre-regulated supply is supplied the 120 Hz rejection is not an issue."
Let me try and understand this....... you were describing the virtues of 120 dB of PSRR, and now you tell us we should use a preregulator as well. The whole point of my discussion of the PSRR was as a matter of comparison with the LM317 which is a pretty good benchmark fromits use in many high end audio line level circuits. It appears to me (and I will be happy to be told otherwise) that your regulator is about 10 times worse in this parameter at 120 Hz, where a large portion of the supply noise in the unregulated supply before the regulator resides.
A lot of people feel that PSRR is one of the most important parameters in a regulator for audio. I have heard the superiority of shunt regulators mentioned by many here. The real advantage of a shunt regulator is very high rejection of noise on the unregulated supply (PSRR) with a reasonable level circuit complexity. I would even dare to speculate that a preregulator would not be required.
"Except, the cable capacity as such does not directly load the regulators, merely indirectly."
Cable reflections (your topic originally, I believe) and RF noise do induce currents from the output of the preamp. These currents come from the supply terminals that the LM 6181 is regulating. Pretty much all of it ;with no decoupling caps for the op amp, (or other circuit types) I would think. RF currents in the output wind up on the supply. The cable capacitance is not directly across the supply but device and layout capacitance and will resonate with the inductive output of the regulator at have frequencies. Audio circuits don't like RF and its presence is often audible.
"While all you say is true you are as usual of poor reading ability, as you seek not to understand but merely to "proove a point". Had you bothered reading yould have noticed in my original post the following remark: PS, attached a rendering of an experimental linestage from a few years ago, drawn by Carlos M. from a description by me, no guarantee for accuracy, I never drew a schema of this, all this is from memory and my memory is notoriously bad."
I think I had a 1k resistor (maybe 10k?) between reference Voltage and Op-Amp input for the reg."
I am sorry I was going by the schematic and did miss that. I still maintain that the LM431 is a much poorer reference the LM329 in terms of noise which is very relevant in audio supplies.
I would not put the LM334 directly in series with the LM431, as both derive there characteristics from feedback circuits, and some pretty nasty circuit stability can result. The transient response of the current sorce and voltage referece ore nasty enough by themselves.
"As said, lowest noise and lowest output impedance as such where not design goals."
Kind of strange, PSRR, noise, and output impedance are the merits on which voltage regulators are judged. There is a pretty wide consensus that they correlate pretty well with how a regulator sounds. The RF stability and rejection are usually addressed with decoupling caps for some pretty good reasons. What is it that this regulator does so well again? No comments on the push pull versus single ended question?
"Funny. I seem to have missed you actually building the circuit. So empiricism seems not to have featured. But semantics is hardly a specialty of mine."
I think you need a new Psychic. You seem to be getting a lot of misinformation on my activities. I don't think I could have come up with all of this from reading data sheets. I have been working with high speed analog circuits for over 15 years. I guess don't understand what your real dispute in my observations and numbers.
I then trust that you must have quite some experience with building and listening to the Jung regulator since you so readily disputed Andy's observations and made statements contrary to my findings as well. From some of your comments, I wonder if you have read the articles about its development. I think technical concerns have been addressed some time ago. The stabilty concerns are kind of funny considering the equally valid stability concerns of the LM6181 circuit Those who live in glass houses..... By the way, I believe Andy knows more about the circuit that even Walt Jung does at this point.
I feel that semantics will become a specialty of yours given the amount of attention you have devoted to it of late. I welcome your input on improving my communication skills.
You stated that you liked the LM6181 regulator sound but switched to valves. Maybe you didn't like it that well then?
Good night from soon to be sleeping giant,
Fred
"As a pre-regulated supply is supplied the 120 Hz rejection is not an issue."
Let me try and understand this....... you were describing the virtues of 120 dB of PSRR, and now you tell us we should use a preregulator as well. The whole point of my discussion of the PSRR was as a matter of comparison with the LM317 which is a pretty good benchmark fromits use in many high end audio line level circuits. It appears to me (and I will be happy to be told otherwise) that your regulator is about 10 times worse in this parameter at 120 Hz, where a large portion of the supply noise in the unregulated supply before the regulator resides.
A lot of people feel that PSRR is one of the most important parameters in a regulator for audio. I have heard the superiority of shunt regulators mentioned by many here. The real advantage of a shunt regulator is very high rejection of noise on the unregulated supply (PSRR) with a reasonable level circuit complexity. I would even dare to speculate that a preregulator would not be required.
"Except, the cable capacity as such does not directly load the regulators, merely indirectly."
Cable reflections (your topic originally, I believe) and RF noise do induce currents from the output of the preamp. These currents come from the supply terminals that the LM 6181 is regulating. Pretty much all of it ;with no decoupling caps for the op amp, (or other circuit types) I would think. RF currents in the output wind up on the supply. The cable capacitance is not directly across the supply but device and layout capacitance and will resonate with the inductive output of the regulator at have frequencies. Audio circuits don't like RF and its presence is often audible.
"While all you say is true you are as usual of poor reading ability, as you seek not to understand but merely to "proove a point". Had you bothered reading yould have noticed in my original post the following remark: PS, attached a rendering of an experimental linestage from a few years ago, drawn by Carlos M. from a description by me, no guarantee for accuracy, I never drew a schema of this, all this is from memory and my memory is notoriously bad."
I think I had a 1k resistor (maybe 10k?) between reference Voltage and Op-Amp input for the reg."
I am sorry I was going by the schematic and did miss that. I still maintain that the LM431 is a much poorer reference the LM329 in terms of noise which is very relevant in audio supplies.
I would not put the LM334 directly in series with the LM431, as both derive there characteristics from feedback circuits, and some pretty nasty circuit stability can result. The transient response of the current sorce and voltage referece ore nasty enough by themselves.
"As said, lowest noise and lowest output impedance as such where not design goals."
Kind of strange, PSRR, noise, and output impedance are the merits on which voltage regulators are judged. There is a pretty wide consensus that they correlate pretty well with how a regulator sounds. The RF stability and rejection are usually addressed with decoupling caps for some pretty good reasons. What is it that this regulator does so well again? No comments on the push pull versus single ended question?
"Funny. I seem to have missed you actually building the circuit. So empiricism seems not to have featured. But semantics is hardly a specialty of mine."
I think you need a new Psychic. You seem to be getting a lot of misinformation on my activities. I don't think I could have come up with all of this from reading data sheets. I have been working with high speed analog circuits for over 15 years. I guess don't understand what your real dispute in my observations and numbers.
I then trust that you must have quite some experience with building and listening to the Jung regulator since you so readily disputed Andy's observations and made statements contrary to my findings as well. From some of your comments, I wonder if you have read the articles about its development. I think technical concerns have been addressed some time ago. The stabilty concerns are kind of funny considering the equally valid stability concerns of the LM6181 circuit Those who live in glass houses..... By the way, I believe Andy knows more about the circuit that even Walt Jung does at this point.
I feel that semantics will become a specialty of yours given the amount of attention you have devoted to it of late. I welcome your input on improving my communication skills.
You stated that you liked the LM6181 regulator sound but switched to valves. Maybe you didn't like it that well then?
Good night from soon to be sleeping giant,
Fred
The Chronicles of Nania
"I have been eyeing this 6181 preamp for quite some time now and I would love to be the guineau pig to build it and report back to this thread. We could even feed some inverted gainclones with it. Thorsten? Batteries or mains supply?"
I would AC couple the output cause the circuit has a BIG DC offset.
With your knowledge of transformers, definitly a lion powered manes supply.
Peace,
Aslan
"I have been eyeing this 6181 preamp for quite some time now and I would love to be the guineau pig to build it and report back to this thread. We could even feed some inverted gainclones with it. Thorsten? Batteries or mains supply?"
I would AC couple the output cause the circuit has a BIG DC offset.
With your knowledge of transformers, definitly a lion powered manes supply.
Peace,
Aslan
A mix seems to work well...
I was playing around with various fixed regulators for a preamp project I've been working on. For various reasons the last few I've tried were On semi's: MC78M15CDT and MC79M15CDT, along with the 12V versions (purchased from Arrow); and National's LM2990S and LM2940CS, also in 12 and 15 V versions (from Digikey). All were TO-263 style.
The scheme is +/- 21VDC (RC filtered: 0.1R with 20kuF per rail) going via 0805 ferrite beads to 15V regs and the output from the 15 regs goes to the 12V regs--15V regs are 'pre-regs' for the 12V regs. (There is circuitry, a load, being driven by the 15V regs and the 12V regs.) Caps on output of regs are Cornell Dubilier SMT Ultra low ESR (SPV107M16M24T; Mouser) 100uF/16V bypassed with 0.1uF PP in 0805 SMT.
The LM series looked pretty clean on my scope, but there were still some high freq garbage, especially on the neg rails. When the MC series were substituted, the neg. rails looked very good, but the pos. rails were absolute garbage. A lightbulb can on, so I swapped out the MC pos. regs and substituted the LM pos. regs. I kept the MC neg. regs in place.
This mixture gave extraordinarily clean rails. While I only have a 20M scope with my lowest scale at 5mV, using a 10x probe showed 5mV of ripple going into the 15V regs and then nothing was visible after this, just a flat line. (To better accentuate matters, I drove the preamp circuit with a 120Hz signal; it syncs my scope better; changing freq didn't reveal any more junk.)
Sonically, the circuit with this mixture of regulators is quieter than I've ever not heard.
To summarize, the best combo: 15V are MC79M15CDT (-15) and LM2940CS (+15) and the 12V are MC79M12CDT (-12) and LM2940CS (+12).
Robert
I was playing around with various fixed regulators for a preamp project I've been working on. For various reasons the last few I've tried were On semi's: MC78M15CDT and MC79M15CDT, along with the 12V versions (purchased from Arrow); and National's LM2990S and LM2940CS, also in 12 and 15 V versions (from Digikey). All were TO-263 style.
The scheme is +/- 21VDC (RC filtered: 0.1R with 20kuF per rail) going via 0805 ferrite beads to 15V regs and the output from the 15 regs goes to the 12V regs--15V regs are 'pre-regs' for the 12V regs. (There is circuitry, a load, being driven by the 15V regs and the 12V regs.) Caps on output of regs are Cornell Dubilier SMT Ultra low ESR (SPV107M16M24T; Mouser) 100uF/16V bypassed with 0.1uF PP in 0805 SMT.
The LM series looked pretty clean on my scope, but there were still some high freq garbage, especially on the neg rails. When the MC series were substituted, the neg. rails looked very good, but the pos. rails were absolute garbage. A lightbulb can on, so I swapped out the MC pos. regs and substituted the LM pos. regs. I kept the MC neg. regs in place.
This mixture gave extraordinarily clean rails. While I only have a 20M scope with my lowest scale at 5mV, using a 10x probe showed 5mV of ripple going into the 15V regs and then nothing was visible after this, just a flat line. (To better accentuate matters, I drove the preamp circuit with a 120Hz signal; it syncs my scope better; changing freq didn't reveal any more junk.)
Sonically, the circuit with this mixture of regulators is quieter than I've ever not heard.
To summarize, the best combo: 15V are MC79M15CDT (-15) and LM2940CS (+15) and the 12V are MC79M12CDT (-12) and LM2940CS (+12).
Robert
Re: Re: local stuff
Well, you got me stumped here. In the beginning post #2 you mentioned you used exactly the same opamp for supply & amp, and told me how to get that working. Then again in post #80 you even capitalize the word EXACTLY. In my book, that's pretty close to advocating it. But it seems we're back to the semantics again.
Maybe Fred and I can get a discount if we sign up for the same communication skills course?😉
Sayonara,
Jan Didden
Kuei Yang Wang said:
Well, you got me stumped here. In the beginning post #2 you mentioned you used exactly the same opamp for supply & amp, and told me how to get that working. Then again in post #80 you even capitalize the word EXACTLY. In my book, that's pretty close to advocating it. But it seems we're back to the semantics again.
Maybe Fred and I can get a discount if we sign up for the same communication skills course?😉
Sayonara,
Jan Didden
he's so modest
Take your compliments where you can get them. I hearby take full responsibily for that belief. I hope that I never implied to anyone that you ever even hinted at such an idea. In self defence, I have read all the Jung, Sulzer, Borbely,Marsh, Didden, Galo, and other's articles in Audio Amateur and Audio Electronics repeatedly. I have also seen your regulator board layout, schematic, and manual as well as your post on the subject of power supplies. If I say anymore you will have to pay me for marketing or write me into your will. (A strategy I use on my nieces as well as the frequently shifting favorite neice status, that my wife gives me the evil eye about. They swear I am their favorite uncle and they will look after me until they can plan my demise without incriminating themselves.)
I will still stand by my assessment, being a licensed Audio Guru of licensed by the state of Texas. In fact to paraphrase Mark Twain's quote on Kipling:
Andy is a stranger to me, but he is a most remarkable man--and I am the other one. Between us, we cover all power supply design knowledge; he knows all that can be known, and I know the rest.
Now get in there and pull your weight for the great regulate debate. (hate to wait, mate) I am exhausted. Was that a sigh of relief I heard ?
Take your compliments where you can get them. I hearby take full responsibily for that belief. I hope that I never implied to anyone that you ever even hinted at such an idea. In self defence, I have read all the Jung, Sulzer, Borbely,Marsh, Didden, Galo, and other's articles in Audio Amateur and Audio Electronics repeatedly. I have also seen your regulator board layout, schematic, and manual as well as your post on the subject of power supplies. If I say anymore you will have to pay me for marketing or write me into your will. (A strategy I use on my nieces as well as the frequently shifting favorite neice status, that my wife gives me the evil eye about. They swear I am their favorite uncle and they will look after me until they can plan my demise without incriminating themselves.)
I will still stand by my assessment, being a licensed Audio Guru of licensed by the state of Texas. In fact to paraphrase Mark Twain's quote on Kipling:
Andy is a stranger to me, but he is a most remarkable man--and I am the other one. Between us, we cover all power supply design knowledge; he knows all that can be known, and I know the rest.
Now get in there and pull your weight for the great regulate debate. (hate to wait, mate) I am exhausted. Was that a sigh of relief I heard ?
- Status
- Not open for further replies.